Enforce cleanup_controller on more exit points of configure_controller#3192
Conversation
Codecov Report❌ Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## master #3192 +/- ##
==========================================
- Coverage 89.12% 89.10% -0.03%
==========================================
Files 158 158
Lines 19587 19601 +14
Branches 1589 1590 +1
==========================================
+ Hits 17457 17465 +8
- Misses 1471 1476 +5
- Partials 659 660 +1
Flags with carried forward coverage won't be shown. Click here to find out more.
🚀 New features to boost your workflow:
|
|
Hi, I have some tests failing due to timing issues, could somebody lend me a hand? |
These are known flaky tests and not related, don't worry. |
Understood, thanks! Anything else I should be doing (eg reaching out to people some channel)? I didn't find any discord or similar in the contribution guide. |
|
There is no need to tag anyone, we will come back when we will have time. You always can help in reviewing one of the many other open PRs ;) |
saikishor
left a comment
There was a problem hiding this comment.
Rest LGTM. Just one minor nit in the tests
99b5dbf to
4499776
Compare
Two of the possible early return paths on controller_manager::configure_controller led to controllers being left in an INACTIVE state instead of rolling back to the UNCONFIGURED state. This specifically affected controllers that were chainable that: - didn't export any state nor reference interfaces - raised an exception while querying their state/reference interfaces Two tests were added to the controller_manager/test_cleanup_controller test suite for these cases.
4499776 to
369c53c
Compare
saikishor
left a comment
There was a problem hiding this comment.
LGTM.
P.S. Do not force push the PRs again, this makes the maintainers to review all the codebase again. Moreover, we will squash the commits
|
Got it, I'll refrain from doing that. What if I want to rebase on top of master though? Do you need anything else from my side to get this merged? Thanks for the feedback! |
You can just sync those commits using the GitHub UI from your fork or from the PR |
|
christophfroehlich
left a comment
There was a problem hiding this comment.
Please fix the failing pre-commit job, I suggest to install pre-commit permanently in this repo.
Two of the possible early return paths on controller_manager::configure_controller led to controllers being left in an INACTIVE state instead of rolling back to the UNCONFIGURED state. This specifically affected controllers that were chainable that:
Two tests were added to the controller_manager/test_cleanup_controller test suite for these cases.
This PR should solve #3153
To send us a pull request, please:
colcon testandpre-commit run(requires you to install pre-commit bypip3 install pre-commit)